Is fish embryo test (FET) according to OECD 236 sensible enough for delivering quality data for effluent risk assessment?

Over the past few years, the fish embryo test (FET) has become widely accepted as an animal-friendly protocol for ecotoxicological research. As Organisation for Economic Co-operation and Development (OECD) 236, the FET has been widely applied for simple mixture exposures under the Registration, Evaluation, Authorisation, and Restriction of Chemicals regulation of the European Union; and now its use is spreading worldwide as a supposedly reliable whole-effluent test (i.e., the testing of complex mixture exposures). However, comparative peer-reviewed data regarding the FET's efficiency for whole-effluent tests are virtually nonexistent. The primary objective of the present study was to make the first comparative test between the FET according to OECD 236 and other standard and slightly modified standard fish protocols used worldwide for whole-effluent tests. For that, we used an untreated hospital effluent considered to be highly toxic but disposed of in municipal sewerage. The base methods were OECD 203 (juvenile), US Environmental Protection Agency method 2000.0 (larvae), and OECD 236 (embryo). We also evaluated the addition of 3 virtually costless sublethal metrics (immobility, nonhatching, and pericardial edema) that could enhance the sensitivity of OECD 236. We observed acute toxicity in all 8 methodologies tested, with a clear escalation in sensitivity (larvae > juvenile ≥ embryo). Larvae were the most sensitive life stage for whole-effluent tests. The addition of sublethal metrics to OECD 236 enhanced its previous sensitivity in over 30%. Thus we conclude that OECD 236 acts below its potential and that the embryonic stage (as used in the FET) may not be the most sensitive life stage for whole-effluent tests. Environ Toxicol Chem 2018;37:2925-2932. © 2018 SETAC.